Enhancing Video Games Policy Based on Least-Squares Continuous Action Policy Iteration: Case Study on StarCraft Brood War and Glest RTS Games and the 8 Queens Board Game

Table 9

Reward function calculations of LSCAPI versus LSPI.

Cases	LSCAPI algorithm			LSPI algorithm
	Used actions	Ratio	Reward	Reward
1	20/24	0.833	27,570	21,500
2	19/24	0.792	30,010	25,000
3	16/24	0.667	33,000	26000
4	21/24	0.875	18,000	13,000
5	23/24	0.958	12,000	8,500
6	22/24	0.917	14,530	11,500
7	20/24	0.833	29,000	22,000
8	22/24	0.917	15,000	8,500
9	21/24	0.875	17,000	14,000
10	22/24	0.917	15,700	13,000
11	22/24	0.917	14,000	12,000